Based on Airline Passenger Satisfaction datasets to training model for classification and prediction passenger's satisfaction on using airline services
Link data: https://www.kaggle.com/datasets/teejmahal20/airline-passenger-satisfaction
- Data preprocessing
1.1. Encode object fields to numeric fields using the Labels encoding method.
1.2. Visualize data to view and find down the way to process
1.3. Fill empty value in dataset by Imputer
1.4. Normalize data by Standard Scaller
1.5. PCAs and select important features
-
Fitting with build-in model
-
Coding our model from scratch
3.1. Logistic Regression
3.2. Multi Layer Perceptrons
3.3. Decision Tree Classifier
3.4. Gaussian Naive Bayes
- Evaluation (f1-score)
4.1. Logistic Regression: 0.87
4.2. Multi Layer Perceptrons: 0.96
4.3. Decision Treee Classifier: 0.94
4.4. Gaussian Naive Bayes: 0.86